Version 0.4.0 Release Notes (spring-summer 2017) 
==============================


New features:

 - New function `XLtable1' exports a manuscript-Table-1-like series of descriptive tables with common structure.
 - This function required some useful changes in "atomic" package functions: `XLoneWay,XLtwoWay' now allow `margins=FALSE' 
 - In `XLregresSummary`, the CI and p-value calculations are now linked, and controlled via the `df` argument. So there are now only 2 options, Normal (default) and T (if `df` is a positive number), vs. unlimited but annoying before. Naturally, the arguments `pfun` and `confac` are now gone.
 - Introduced `niceRound', a proper-rounding utility useful beyond this package.
 - `XLtwoWay' now reports Chi-squared p-values as default. User can provide another test function, or suppress the test.
 - Similarly for `XLoneWay`, except here the default is no test, b/c wanting to p-value a one-way table is far less common.
 - Added Yihui's "roxygenize hack" to DESCRIPTION, to enable full help files when installing from GitHub. Not included in CRAN version, per CRAN requirements.


Version 0.3.2 Release Notes (summer 2016) 
==============================

Bug fixes:

- Added `iqrString` to the the exported functions

New features:

- `XLtwoWay` can now do the percents in parentheses next to counts, like `XLoneWay`; actually it's the default. You can still also do side-by-side, or drop the percents.
- All(?) rounding is now wrapped by `format()` to make sure the actual requested number of digits is showing, rather than R/Excel's annoying "feature" of rounding it further for you in case your number happens to be roundish.
- In `XLregresSummary`, a p-value previously rounded down to zero, will now show as "<p_0", with p_0 being the smallest positive number one could round to (default: "<0.001" because `pround=3`).
- In `XLoneWay` and `XLtwoWay`, I bumped the default n value for switching between rounding to whole percents or leaving one decimal digit, from 200 to 500 based on experience. Of course, you can control the rounding directly via `digits`.


Version 0.3.1 Release Notes (early summer 2014) 
==============================

Bug fixes:

- Added automatic file-saving to `XLgeneric` (as promised by help text)
- `row1,col1` are now "as advertised", i.e. the export really begins at the cell you tell it to (in some functions it was only "approximate").
- Removed some outdated text from help pages.

New features:

- Added `XLaddText` for putting a text string anywhere in the file. This is now also the behind-the-scenes method for adding titles to tables in all other export functions.
- `XLregresSummary` now also accepts a matrix as input, instead of providing effect names, point estimates and SEs separately (which is still possible). 
- in `XLgeneric`, you can now modify the column names as well as the row names.


Version 0.3 Release Notes (spring 2014, targeting R 3.1.0) 
==============================

- Added `XLgeneric` for exporting a generic rectangular dataset
- Fixed bug on handling missing values in XLoneWay, XLtwoWay, and explicitly added 'useNA' option passed directly on to table().
- Added option not to write a percents table in XLtwoWay.
- Fixed bug in XLoneWay that returned counts as text when `combine=FALSE`.
- Added an `XLConnect::saveWorkbook` command to all export functions, so that all changes are immediately saved to file.
- Added this NEWS file...


Version 0.2 (October 2013): first release   
=====================
Hello World!
